BACKGROUND Diabetic kidney disease (DKD), characterized by increased urinary microalbumin levels and decreased renal function, is the primary cause of end-stage renal disease. Its pathological mechanisms are complicated and multifactorial; Therefore, sensitive and specific biomarkers are needed. Urinary exosome originate from diverse renal cells in nephron segments and partially mirror the pathological changes in the kidney. The microRNAs (miRNAs) in urinary exosome are remarkably stable and highly tissue-specific for the kidney. AIM To determine if urinary exosomal miRNAs from diabetic patients can serve as noninvasive biomarkers for early DKD diagnosis. METHODS Type 2 diabetic mellitus (T2DM) patients were recruited from the Second Hospital of Hebei Medical University and were divided into two groups: DM, diabetic patients without albuminuria [urinary albumin to creatinine ratio (UACR) < 30 mg/g] and DKD, diabetic patients with albuminuria (UACR >= 30 mg/g). Healthy subjects were the normal control (NC) group. Urinary exosomal miR-145-5p, miR-27a-3p, and miR-29c-3p, were detected using real-time quantitative polymerase chain reaction. The correlation between exosomal miRNAs and the clinical indexes was evaluated. The diagnostic values of exosomal miR-145-5p and miR-27a-3p in DKD were determined using receiver operating characteristic (ROC) analysis. Biological functions of miR-145-5p were investigated by performing Gene Ontology analysis and Kyoto Encyclopedia of Genes and Genomes pathway enrichment. RESULTS Urinary exosomal expression of miR-145-5p and miR-27a-3p was more upregulated in the DKD group than in the DM group (miR-145-5p: 4.54 +/- 1.45 vs 1.95 +/- 0.93, P < 0.001; miR-27a-3p: 2.33 +/- 0.79 vs 1.71 +/- 0.76, P < 0.05) and the NC group (miR-145-5p: 4.54 +/- 1.45 vs 1.55 +/- 0.83, P < 0.001; miR-27a-3p: 2.33 +/- 0.79 vs 1.10 +/- 0.51, P < 0.001). The exosomal miR-145-5p and miR-27a-3p positively correlated with albuminuria and serum creatinine and negatively correlated with the estimated glomerular filtration rate. miR-27a-3p was also closely related to blood glucose, glycosylated hemoglobin A1c, and low-density lipoprotein cholesterol. ROC analysis revealed that miR-145-5p had a better area under the curve of 0.88 [95% confidence interval (CI): 0.784-0.985, P < 0.0001] in diagnosing DKD than miR-27a-3p with 0.71 (95%CI: 0.547-0.871, P = 0.0239). Bioinformatics analysis revealed that the target genes of miR-145-5p were located in the actin filament, cytoskeleton, and extracellular exosome and were involved in the pathological processes of DKD, including apoptosis, inflammation, and fibrosis.<br /> CONCLUSION Urinary exosomal miR-145-5p and miR-27a-3p may serve as novel noninvasive diagnostic biomarkers or promising therapeutic targets for DKD.